The difficulty for Antarctica is regulatory, not ease.
Returning to the difficulty of agriculture on Antarctica …
I happened to encounter this article a week ago.
It’d be darn difficult to farm there, but it’s getting easier.
Yes, it would be ideal. At risk of hijack, I think the fact that Elon Musk has no plan to operate an Antarctic habitat is evidence that he’s also not planning a Mars habitat.
That misses the point that a Mars colony is not on Earth. Specifically it’s a ~260 day journey away from Earth– and any totalitarians with nukes.
If Earth dies, Mars dies.
You just explained the entire reason why the prototype colony has to be built on Earth. And if it’s not getting built on Earth, then it’s not getting built anywhere.
Antarctica is easy mode for an off-planet colony. If you can’t do it there, you’re certainly not going to do it on Mars.
The whole point is to make it self-sustaining. That’s probably a century off. But it’s a century off from whenever we start. So we may as well start now.
As this whole thread has gone into, Antarctica is actually poorer in some things than Mars is.
It’s still a lot easier overall than Mars.
Or if you want to go really easy mode, then build a sealed, self-sustaining colony in Kansas. But nobody’s managed that yet, either. Nobody’s even seriously trying.
And no one is saying that we’re Musk’s going to just plop 50,000 people down on Mars and if something goes wrong then they all die, too bad. No settlement on Mars is going to expand past its proven capacity to support, and if that support doesn’t happen the expansion doesn’t happen. Plus, no one is saying that perfect recycling and 100% self-sufficiency must happen right at the start. In-between a mere base completely reliant on supplies and equipment from Earth and an independent self-sufficient civilization on Mars, there are any number of degrees of partial sufficiency adequate to support N number of people. I guarantee you that a founding precept of a Martian colony is going to be reserves of liquid oxygen, potable water and freeze-dried food sufficient to sustain everyone for a full Earth-Mars synodic period, until either resupply or evacuation can happen.
It is interesting that so many folks have jumped to this as rehearsal for Moon or Mars colonies when the OP doesn’t seem to have wanted anything more than US 1800s frontier level survival in a hostile Earth climate.
Sort of a pre-contact Inuit settlement but initially staffed by 21st Century folks.
The problem is that right now, any Mars settlement’s proven capacity to support is 0. And the way you prove that number higher starts with Earth settlements.
There really isn’t, though. It’s too expensive and time-consuming to get material to Mars. The viable levels are nothing at all, a Mars base that requires resupply of only a small amount of lightweight “vitamin” supplies like computer chips and diamond tool bits, and a completely independent Mars base. Anything in between no base at all and a base that imports only “vitamins” is prohibitively expensive and unreliable. That’s the reason why Musk is pushing for “colonists” who never return as the first humans to walk on Mars.
Maybe a Mars colony will have a multi-year’s supply of food, water, and air before humans ever go there, but if that’s the case, it’ll be because robots produced all of those things from in-situ resources. It’s just not possible otherwise.
Eh. Water recycling is not hard. It’s done on the ISS already. The main consumable aside from that is food, which will be the first order of business to advance the state of a colony. But suppose that proves to be more difficult than anticipated. A human needs about 1 kg/day of food (assuming judicious use of dehydration). And Starship will cost about $2000/kg to Mars, even somewhat pessimistically (optimistically it will be <$1000). So a lifetime (50-year) supply of food for one person costs about $36M. So supplying a 100-person colony for a lifetime is basically pocket change for Musk. Each Starship can land a lifetime supply of Spam and dehydrated fruit for about 5 people.
Everything else is basically in the noise–the solar panels will degrade over time, but it takes so long (decades) that the consumption rate is tiny. Same for any other industrial equipment. Just look at how much stuff the item produces over its lifetime. It’s probably >100x for most things, whether oxygen generators or 3D printers. Oh, and they get a free hundred tons of high-grade metal with every food delivery.
A million people can’t be economically supplied with food this way, so we can presume the 100-person colony won’t grow further until they solve the food problem (at least mostly). But it can stay in that state for a very long time, and in the meantime the colonists will have a strong incentive to solving it.
If we postulate that the indefinite “budget” for a Mars colony is 1000-5000 metric tonnes of supplies every synodic launch window (~780 days), then what that cargo is and how many people it can support depends on how successful in-situ resource utilization, food production and recycling efficiency is. At a minimum there will be no colonization if liquid oxygen can’t be produced on the industrial scale necessary to refuel Starships for return to Earth; so let’s say that breathing is a given, minus CO2 scrubbing and equipment maintenance/replacement. ETA: and producing the liquid methane for the return trips means obtaining some regular supply of hydrogen, probably water ice, in-situ. So if there’s a colony at all they’re already producing their own air and water open-cycle. If they can recycle previously produced air and water all the better.
I anticipate that most of the Starship trips will be one-way. There won’t be any return trips on the first synod. Maybe by the second there will enough to start returning a ship every synod. That’s enough to bring anyone back that wants to leave. It still requires a very significant industrial plant with megawatts of power… but that’s really just not a huge challenge. You could deploy a 10 MW solar field with a handful of people in weeks. The lack of weather means you can just plop down lightweight frames with the panels attached. Or even just unroll them directly on the ground.
Human O2 requirements can be handled by CO2 electrolysis, which has already been demoed on Mars with the MOXIE experiment. But for propellant production you probably want to use ice as the feedstock. As you note, you need it anyway for the hydrogen.
To my unschooled eye, the limfac is reliability & spare parts for all the must-work infrastructure. It’s great to mine ice for the electrolysis machines. As long as the ice mining machines don’t break down. Nor the electrolysis machines. Nor do the people who’re trained in their repair breaking down with appendicitis, gizzard cancer, or broken legs after a fall off the mining machinery.
Yes, but you can fall at almost 3 times the height with no issue 
.
I think in terms of basic equipment, they’ll just dump tons and tons of it alongside everything else. This isn’t a NASA-style effort where they try to shave off every gram. There’s no need to pack light.
I know you don’t watch much TV, but you might (somewhat) appreciate For All Mankind, an alternate history of the space race starting in a slightly different 1969 and moving forward a few decades. Where if there is a stupid way to screw up and get people killed, they are always sure to find it.
There’s nothing you could extract on Mars that you couldn’t extract in Antarctica, unless easier-available H2O2 was a major concern for you.
The one thing that Mars is richer in, iron, isn’t exactly a scarce commodity in Antarctica.
Sorta. I totally see and agree with your point that the cost/kg numbers using Starship very much change the equation.
But if the goal is to support a real colony, then any serious over-provisioning of [whatever] represents a lost opportunity to carry across space and support on the surface the real reason you’re there: more people.
In Earth history many colonies have collapsed when climate or locusts or whatever suddenly greatly reduced the local carrying capacity and what had been a happily sustainable population was suddenly 10x too many. Sucks to be 90% of those people.
So you don’t want to design your Mars colony so they’re one small failure away from losing people to starvation, freezing, or asphyxiation, but you also don’t want to send equipment enough for a thousand person settlement then send 15 people to live in it. Unless the idea is the 15 are supposed to birth the other 985. 
[Obligatory Dr. Strangelove clip incorporated by reference.]